Article Text
Abstract
Background Susac syndrome is characterised by the triad of encephalopathy with or without focal neurological signs, branch retinal artery occlusions and hearing loss. Establishment of the diagnosis is often delayed because the triad is complete only in a minority of patients at disease onset. This leads to a critical delay in the initiation of appropriate treatment. Our objective was to establish criteria for diagnosis of either definite or probable Susac syndrome.
Method The establishment of diagnostic criteria was based on the following three steps: (1) Definition of a reference group of 32 patients with an unambiguous diagnosis of Susac syndrome as assessed by all interdisciplinary experts of the European Susac Consortium (EuSaC) team (EuSaC cohort); (2) selection of diagnostic criteria, based on common clinical and paraclinical findings in the EuSaC cohort and on a review of the literature; and (3) validation of the proposed criteria in the previously published cohort of all Susac cases reported until 2012.
Results Integrating the clinical presentation and paraclinical findings, we propose formal criteria and recommend a diagnostic workup to facilitate the diagnosis of Susac syndrome. More than 90% of the cases in the literature fulfilled the proposed criteria for probable or definite Susac syndrome. We surmise that more patients could have been diagnosed with the recommended diagnostic workup.
Conclusions We propose diagnostic criteria for Susac syndrome that may help both experts and physicians not familiar with Susac syndrome to make a correct diagnosis and to prevent delayed treatment initiation.
Statistics from Altmetric.com
Introduction
Susac syndrome (SuS) is a rare and presumably autoimmune mediated disease affecting the small arterial vessels of the brain, inner ear and retina.1–3 The clinical triad comprising brain involvement with encephalopathy, focal neurological deficits and headache, hearing loss and branch retinal artery occlusions (BRAOs) is considered pathognomonic. However, the individual manifestations may develop only as single events over a prolonged disease course, may be too subtle to be noticed or may be masked by severe encephalopathy. Consequently, the complete triad is present in only 13% of patients at disease onset.1 The lack of diagnostic criteria often leads to a delayed or wrong diagnosis and a critical protraction of an appropriate treatment.4 ,5 Of particular note, immune treatment for multiple sclerosis (MS), which is the most common misdiagnosis of SuS,6 may lead to exacerbation of the disease.7
In the current study, we, the members of the European Susac Consortium (EuSaC, http://www.eusac.net), propose easily applicable diagnostic criteria to be used by experts and physicians not familiar with SuS in order to facilitate a timely and reliable establishment of the diagnosis as a basis for treatment decisions and for the use in upcoming trials.
Methods
Protocol approval, registration, and patient consents
The study was approved by the local ethics committee of the University Hospital Münster, Germany (2013-330-f-S, 2011-305-f-S und 2010-223-f-S). All participants gave written informed consent.
Development of diagnostic criteria
The establishment of diagnostic criteria was based on three steps:
Definition of a reference cohort comprising patients with an unambiguous diagnosis of SuS, according to expert consensus (EuSaC cohort).
Selection of diagnostic criteria, based on common clinical and paraclinical findings in the EuSaC reference cohort and on a thorough review of the current literature.
Validation of the proposed criteria in the previously published cohort of all SuS cases reported until 2012.1
Definition of a SuS reference cohort (EuSaC cohort)
Using the EuSaC database, a reference cohort comprising patients with SuS with an unambiguous diagnosis was defined. For this purpose, we reviewed the medical records of all patients with SuS referred to the Susac centre in Münster, Germany, between 2004 and 2014. Clinical and paraclinical data were assessed by all members of the interdisciplinary SuS team, and only patients in whom all the EuSaC experts ascertained that the diagnosis of SuS unequivocally received a definite diagnosis of SuS were included.
Selection of diagnostic criteria
A systematic search of electronic databases (PubMed, EmBase) was performed to identify relevant data on clinical and paraclinical characteristics of SuS from 1990 to 2016 using the text terms: ‘Susac syndrome’, ‘Susac's syndrome’, ‘retinocochleocerebral vasculopathy’ and ‘SICRET syndrome’ (small infarction of cochlear, retinal and encephalic tissue syndrome).
After harmonising the clinical and diagnostic features of the EuSaC cohort (part 1) and the information provided by the literature search (part 2), a set of diagnostic criteria and categories was chosen by our group in a consensus decision.
Application of diagnostic criteria to published SuS cases
After developing diagnostic criteria, we applied the proposed criteria to cases reported in the literature until 2012 (based on the detailed patients' data provided by reference1).
From all previously published 304 patients with SuS,1 192 patients in whom results of brain, ear and eye examinations were reported or referred to as ‘not assessed’ were considered as sufficiently documented cases and were included in the analysis. Cases without mention of a diagnostic workup of one of the three organs were excluded from the analysis.
Results
Reference cohort of patients with Susac
Thirty-two patients of the EuSaC database with an unequivocal diagnosis of SuS as defined above were selected for the EuSaC cohort and served as a gold standard. The female to male ratio was 2.2:1. The mean age at disease onset was 30.5 (±9.6) years. All patients had an ophthalmological workup including fluorescein angiography (FLA) and an examination of the inner ear function, including vestibular function tests. In 31/32 patients, advanced MRI, including diffusion tensor imaging (DTI) and high-resolution three-dimensional T1-weighted sequences, had been performed. Cerebrospinal fluid (CSF) data were available from 29/32 patients. Individual clinical and paraclinical data of the EuSaC cohort patients such as age, sex, age at onset, disease duration, initial and subsequent clinical features, numbers of exacerbations, response to treatment, cranial MRI data, ophthalmological investigations including retinal FLA and visual field perimetry, audiometry, and serological and CSF data are displayed in table 1.
Proposed diagnostic criteria
Our proposed criteria for the diagnosis of SuS are summarised in box 1.
Overview of the diagnostic criteria for SuS
Definite diagnosis of SuS:
Each criterion (1; 2; 3) with subcriteria (i; ii) has to be met.
Brain involvement
Symptoms and clinical findings: New cognitive impairment and/or behavioural changes and/or new focal neurological symptoms and/or new headache.*
Imaging: Typical findings on cranial MRI—hyperintense, multifocal, round small lesions, at least one of them in the corpus callosum (‘snowball’) in T2 (or FLAIR) weighted sequences.†
To fulfil 1, at least one of the clinical findings and the typical MRI findings have to be documented.
Retinal involvement
Clinical findings and symptoms not required.‡
Ophthalmological examination: BRAOs or AWH in fluorescein angiography or characteristic signs of retinal branch ischaemia in funduscopy or SD-OCT.
To fulfil 2, at least one BRAO or AWH in fluorescein angiography or retinal branch ischaemia in funduscopy or corresponding damage in SD-OCT has to be documented.
Vestibulocochlear involvement
Symptoms and clinical findings: New tinnitus and/or hearing loss and/or peripheral vertigo.
Examination of inner ear function: Hearing loss must be supported by an audiogram;§ vestibular vertigo must be supported by specific diagnostics.¶
To fulfil 3, at least one of the clinical findings must be present and hearing loss or vestibular vertigo must be supported by specific investigations of the inner ear function.
Probable SuS:
Incomplete triad, only two of the three aforementioned criteria 1–3 are fulfilled.**
Patients in which SuS is not the most probable diagnosis (‘possible’ SuS):
In every other patient showing some clinical and/or paraclinical findings of the above triad, but not fulfilling I or II, SuS has to be included in the differential diagnoses, but should not be considered as the most probable diagnosis.Each criterion is divided into two subcriteria: symptoms and clinical findings; and apparative diagnostics.
*Headache can be considered as a symptom of brain involvement if it is new and precedes the other symptoms by not more than 6 months. Headache must be described as migrainous or oppressive, not as a trigemino-autonomous headache.
†MRI T1 lesions can only support the diagnosis if the lesions are: well-demarcated hypointense; affecting the grey matter; Gd enhancing.
Leptomeningeal Gd enhancement on T1-weighted images also supports the diagnosis. Clinical brain involvement without typical lesions on MRI or MRI lesions without clinical signs of brain involvement is not sufficient for the diagnosis.
‡Clinical findings and symptoms are not mandatory because BRAO may be silent.
§For hearing loss: mainly ‘sensorineural hearing loss’, low or mid-tone frequencies and pantonal.
¶For vertigo: caloric testing of the vestibular organ and vestibular evoked myogenic responses.
**Example: patient with headache, tingling of the upper limbs, snowball lesions of the corpus callosum on MRI, as well as tinnitus, and hearing loss in the audiogram, but no BRAO, and exclusion of other diseases. These patients should be considered as having probable SuS and should be thoroughly and frequently examined. Treatment should be considered as in definite SuS.
Exclusion criteria are not defined because exclusion of other diseases is a necessary precondition.
AWH, arterial wall hyperfluorescence; BRAOs, branch retinal artery occlusions; FLAIR, fluid-attenuated inversion recovery; Gd, gadolinium; SD-OCT, spectral domain optical coherence tomography; SuS, Susac syndrome.
Applying these criteria, two categories of diagnostic accuracy were established:
Definite SuS: Patients with an unequivocal clinical and/or paraclinical involvement of all three main organs (ie, fulfilling the typical clinical triad).
Probable SuS: Patients with an unequivocal clinical and/or paraclinical involvement of two of the three main organs.
In patients with some of the typical features of SuS, in whom SuS could not be established as the most probable diagnosis because they do not fulfil even two of the necessary items, the diagnosis of SuS should be considered as possible. These patients need careful and frequent follow-up since the diagnosis could evolve in the future.
In the following paragraphs, the diagnostic items of the three organ manifestations are further specified.
Brain involvement
The patient has to exhibit symptoms and clinical signs suggestive of brain involvement, that is, alteration of consciousness or new cognitive impairment or behavioural changes, and/or new focal neurological symptoms, and/or new headache. Headache has a high prevalence in the normal population, and its symptomatology differs widely in patients with SuS. However, headache is a very common symptom in SuS. We therefore consider headache only relevant for the diagnosis of SuS if it is new, described as migrainous or oppressive, and precedes other manifestations by not more than 6 months.
Additionally, characteristic neuroimaging findings, defined as T2/fluid-attenuated inversion recovery hyperintense multifocal, roundish brain lesions, with at least one of them centrally located in the corpus callosum (‘snowball’, figure 1) are mandatory.8 Variable sizes and shapes, including rectangular, triangular or linear lesions (‘spokes’ and ‘icicles’), can sometimes be observed.9 Sharply confined T1-hypointense lesions with or without gadolinium enhancement, grey matter lesions and leptomeningeal gadolinium enhancement, all support the diagnosis but do not count for themselves.8 Definite brain involvement is defined as the presence of at least one of the aforementioned clinical findings plus characteristic MRI findings.
Retinal involvement
To meet the ‘retinal involvement criterion’, at least one of the following findings should be documented in the ophthalmological examination: (1) at least one acute BRAO in FLA or funduscopy, (2) focal segmental staining of the arterial wall (arterial wall hyperfluorescence, AWH) near the site of obstruction (figure 2) in FLA,10 or (3) sectorial damage of the inner retinal layers from the retinal nerve fibre layer through to the outer plexiform layer in optical coherence tomography (OCT).11 AWH in patients with SuS indicates the impaired integrity of the retinal vessel wall, has to be differentiated from fluorescein leakage, can be considered as a marker for acute disease activity and has not been demonstrated in other retinal vasculitides.10 Visual field deficits or scintillating scotoma may support the diagnosis and help to initiate the necessary diagnostic procedures, but are not sufficient to fulfil the retinal involvement criterion. In particular, peripheral retinal arterial wall plaques are typical for SuS, but they are often transient.2 Spectral domain OCT (SD-OCT) recently turned out to reliably detect highly disease-specific retinal pathologies, rendering it a seminal tool for retinal involvement analyses in later disease stages, when there may be no more BRAO and AWH detectable by FLA.11 ,12
Vestibulocochlear involvement
To qualify for the ‘vestibulocochlear involvement criterion’, at least one of the following clinical manifestations should be present: new (1) tinnitus, (2) hearing loss or (3) peripheral vertigo. Since these clinical symptoms are rather unspecific, hearing loss should be objectified by a pure-tone or speech audiogram, favourably for low or mid-tone frequencies (figure 3). Peripheral vertigo should be supported by caloric testing of the vestibular organ, nystagmography and/or vestibular evoked myogenic potentials.
Rejected items
Although the following features might be helpful and supportive for the diagnosis of SuS, they were considered too unspecific to be used as diagnostic criteria.
Histology
Brain biopsies reported in the literature revealed focal microangiopathic and/or gliotic changes in 28 of 35 patients.1 Subtle perivascular and/or leptomeningeal infiltrates and arteriolar wall thickening have been reported, but some samples were reported to be completely normal.1 A recently published report of brain biopsies in three patients described microinfarcts, thickened hyalinised vessel walls due to collagen deposition and patchy infiltration with CD8+ T cells as well as perivascular lymphocytic cuffing in the leptomeninges, with enlarged and reactive endothelial cells leading to near occlusion of the vessel.13
Findings from brain, muscle and skin biopsies1 ,13 ,14 are vastly unspecific, differ widely and may rather be helpful to exclude other diseases. In our opinion, the diagnosis of SuS can be established with less-invasive techniques, and biopsy findings need not necessarily be supportive, thus leading to false-negative results.
Laboratory markers
Specific serological markers for SuS do not exist. Antinuclear autoantibodies have been described in patients with SuS, but do not occur more often than in healthy controls.1 Clotting abnormalities have also been reported, but the low number of cases does not support a causative relation.1 A role for antiendothelial cell antibodies (AECAs) in SuS has been repeatedly discussed.1 ,15 Although mean AECA titres were significantly higher in patients with SuS than in healthy controls, recent studies showed that AECA titres >1:100 were found in only 25% of patients with SuS.15 Since various types of antibodies other than AECA have been described in 26% of patients with SuS, these findings are rather not sensitive.15
Routine CSF examination often shows an unspecific elevation of total protein, indicating a disruption of the blood–brain barrier. A mild pleocytosis with a median cell count of 12 cells/µL is present in 45% of SuS cases,1 whereas intrathecal IgG synthesis or oligoclonal bands are rarely found.1
In summary, a thorough workup of serum and CSF is necessary to exclude differential diagnoses but CSF findings currently do not qualify for SuS diagnostic criteria.
Therapeutic response
Prospective or randomised controlled treatment trials for SuS do not exist due to its rarity and often delayed diagnosis. Treatment is empirical and based on the hypothesis of an autoimmune inflammatory endotheliopathic aetiology1 ,15 and some similarity to juvenile dermatomyositis.1 ,4 ,5 In the acute episode, treatment comprises high-dose corticosteroids. To avoid further disease attacks or progression, a continuous immunosuppressive therapy is currently recommended.1 ,4 ,15 ,16 Response to immune treatment was not chosen for the criteria as an 'ex juvantibus' prove of the diagnosis. This is due to the difficulty to define a clinical improvement in patients with SuS and a lack of specificity since some important differential diagnoses of SuS would also improve with immune therapy, for example, vasculitis or MS.
Other MRI analysis techniques
Although showing a very characteristic pattern in voxel-based group-level analyses,17 DTI was not chosen as a diagnostic criterion because the elaborated method is restricted to specialised centres. The same limitation holds true for the innovative technique of ultrahigh field MRI at 7 Tesla (7T).18 Furthermore, the diagnostic impact of DTI and 7T MRI at a single-subject level still has to be proven. Diffusion-weighted MRI with apparent diffusion coefficient values results in heterogeneous findings and was therefore not included in the criteria. If present, these abnormalities support the ischaemic origin of the lesions.19
Application of diagnostic criteria in patients described in the literature
Among the 192 published cases with SuS and sufficient data published by 2012,1 107 patients (56%) met the proposed criteria for ‘definite SuS’. Furthermore, 68 of 192 (35%) patients fulfilled the criteria for ‘probable SuS’.
About 127/192 (66%) patients met both clinical and MRI subcriteria for the brain involvement criterion. While almost all patients (179/192 cases; 93%) showed clinical evidence of central nervous system (CNS) involvement, brain MRI was either not performed at all, or the results were not reported in 61/62 cases (98%). The incidence of the clinical symptoms (headache/encephalopathy/focal neurological symptoms) were almost equally distributed among patients (137/147/132 cases; 71%/77%/69%).
The requirements for the retinal involvement criterion were fulfilled in 185 of 192 patients (96%). In five of the seven negative cases (71%), FLA or funduscopy was not performed. Remarkably, only 54 patients (28%) reported visual impairment, but 111 (58%) had visual field defects in ophthalmological investigations. Clinical signs or symptoms were therefore not considered mandatory in patients with paraclinical evidence of BRAO as outlined above. SD-OCT results of the 192 patients with SuS were not reported.
The clinical and paraclinical requirements for vestibulocochlear involvement were met in 162/192 reported cases (84%). Vertigo or tinnitus was not specified in the literature. Thus, all reported cases that met the item had hypoacusis proven by an audiogram. Within this group, 119 (62%) suffered from bilateral hypoacusis, while unilateral hypoacusis was reported in only 43 cases (22%). Only six cases (3%) had a normal audiogram. In 24 cases (13%) that formally did not meet the criteria of vestibulocochlear involvement, no examination of inner ear function was performed.
Among patients fulfilling the definition for probable SuS, 50/68 (74%) met the criteria of ‘vestibulocochlear involvement’ and ‘retinal involvement’, but not the ‘brain involvement’ criteria. Of note, in almost all of the cases (48/50 patients), either brain MRI was not performed or the results were not reported. Fifteen cases (8%) met the ‘retinal involvement’ and ‘brain involvement’ subitems, but not the ‘vestibulocochlear involvement’ item. Of these, 13 cases (7%) had no examination of the inner ear function. Three patients (2%) fulfilled the ‘brain involvement’ and ‘vestibulocochlear involvement’ criteria but not the ‘retinal involvement’ criteria, while two (1%) had had no FLA or funduscopy at all.
In most cases (11 of the 17 patients; 65%) meeting only one criterion, a branch retinal arteriopathy was detected, but neither the criteria for vestibulocochlear involvement nor those for ‘brain involvement’ were met. None of these 11 patients had been investigated by brain MRI, and in eight cases (47%) no examination of the inner ear function had been performed. Two patients (1%) with either vestibulocochlear involvement or ‘brain involvement’ did not show any involvement of the two other organs. Both patients lacked cranial MRI, and in only one of them FLA or funduscopy was performed. Similarly, in two cases (1%) with isolated CNS manifestations, none had had an examination of the retinal arteries, and one patient was not subjected to examination of the inner ear function.
Discussion
In a patient presenting with encephalopathic symptoms, focal neurological signs and headache of recent onset, branch retinal arteriopathy and vestibulocochlear dysfunction in combination with typical MRI or FLA findings, the diagnosis of SuS is not difficult to establish, and differential diagnoses are sparse. However, at initial presentation, the triad is only occasionally complete,1 and making an early diagnosis before completion of the triad can become challenging. Nevertheless, early diagnosis is essential for appropriate treatment of the disease to prevent severe damage. We hypothesised that standardised and easily applicable diagnostic criteria for SuS would accelerate the establishment of the correct diagnosis. Therefore, we propose diagnostic criteria that are easy to apply in clinical daily routine.
More than 90% of the SuS cases reported in the literature of the cohort until 20121 fulfilled our proposed criteria for definite or probable SuS, with 56% meeting the criteria of definite SuS. The number of definite and even probable cases would most likely have been significantly higher if all patients had received the full diagnostic workup recommended for suspected SuS. This is probably due to the fact that some of the case reports were focused on the ophthalmological or vestibulocochlear manifestations.
In an interdisciplinary approach involving neurologists, (neuro)radiologists, ophthalmologists and otorhinolaryngologists, all experienced in SuS, numerous features of the disease have been evaluated with respect to eligibility for diagnostic criteria.
A strength of our study is the large numbers of patients in both cohorts (literature and EuSaC cohort), considering that SuS is an orphan disease. Another strength of this work is that the team of experts consisted of experienced physicians of all the involved disciplines (neurology, neuroradiology, ophthalmology, otorhinolaryngology, neuropsychology and epidemiology). Limitations are the incomplete data sets of some patients in the literature, the lack of a known gold standard for the diagnosis of SuS to compare our criteria with, and the retrospective approach.
Conclusion
In this study, we present the first diagnostic criteria for SuS, based on clinical and paraclinical manifestations in the commonly affected organ systems: brain, inner ear and retina. To establish the diagnosis, every patient with suspected SuS should receive a neurological examination, a brain MRI, an ophthalmological examination including FLA, and an examination of the inner ear function including an audiogram and functional testings of the vestibular organ. The aim of the proposed diagnostic criteria for SuS is to prevent misdiagnosis and delayed treatment, which could result in irreversible damage of the affected organs. A premature aggressive immune treatment, on the other hand, could lead to disproportionate side effects. The proposed diagnostic criteria may help in the decision whether to treat aggressively or to rather perform watchful waiting. We recommend a specific treatment in patients with a definite or a probable diagnosis of SuS.
References
Footnotes
IK, JD and MR contributed equally.
Contributors The authors belong to the European Susac Consortium (EuSaC) and the EuSaC study group. Every author contributed to the collection of data, the conception of the project, the design of the study and the acquisition and analysis of data. IK, JD, MR, YB, BS, AA and TD drafted the manuscript and provided a significant portion of the figures.
Competing interests IK received travel support and speaker honoraria from CSL Behring and receives a research grant (IMF grant ‘The pathophysiological role and clinical impact of the immune system in Susac syndrome’ (KL111421)). JD received research support from Novartis and Bayer Healthcare, speaker honoraria from Novartis, Teva, Genzyme, Biogen, Allergan, Merck-Serono, and Bayer Healthcare, honoraria for advisory from Teva, Genzyme, Novartis, and Bayer Healthcare, travel support from Bayer Healthcare, Biogen, and Novartis. MR received speaker honoraria from Novartis and Bayer Vital GmbH and travel reimbursement from Bayer Schering, Genzyme and Biogen Idec. CCG's work has been funded by the German Research Foundation individual research grant ‘The role of natural killer cells in the immunoregulation of multiple sclerosis’ (DFG, GR3946/2-1) and the IMF grant ‘The pathophysiological role and clinical impact of the immune system in Susac syndrome’ (KL111421). CCG received speaker honoraria and travel expenses for attending meetings from Genzyme, Novartis Pharma GmbH, and Bayer Healthcare. RG has received research grants from Novartis, speaker honoraria from Novartis and Bayer healthcare and travel reimbursement from Bayer healthcare. FP was supported by German Research Foundation (DFG exc257), BMBF Competence Network Multiple Sclerosis, EU FP7 Framework Program (combims.eu), Guthy Jackson Charitable Foundation, Arthur Arnstein Foundation; research support and personal compensation for activities with Alexion, Biogen, Chugai, Teva, Genzyme, Merckserono, Novartis, Bayer and Medimmune. HWi is member of Scientific Advisory Boards/Steering Committees for Bayer Healthcare, Biogen Idec, Sanofi Genzyme, Merck Serono, Novartis, Roche and Teva. He received speaker honoraria and travel support from Bayer Vital GmbH, Bayer Schering AG, Biogen, CSL Behring, EMD Serono, Fresenius Medical Care, Genzyme, Merck Serono, Omniamed, Novartis and Sanofi Aventis and Teva. He received compensation as a consultant from Biogen Idec, Merck Serono, Novartis, Omniamed, Roche and Sanofi Genzyme. He has got research supports from Bayer Healthcare, Bayer Vital, Biogen Idec, Merck Serono, Novartis, Sanofi Genzyme, Sanofi US and Teva Pharma as well as German Ministry for Education and Research (BMBF), German Research Foundation (DFG), Else Kröner Fresenius Foundation, Fresenius Foundation, Hertie Foundation, Merck Serono, Novartis, NRW Ministry of Education and Research, Interdisciplinary Center for Clinical Studies (IZKF) Muenster, RE Children's Foundation. TD received honoraria and travel expenses from Genzyme, Shire, Bristol-Myers Squibb, Boehringer-Ingelheim Pharma, Sanofi Aventis, Eisai, Novartis, Bayer Vital, Merz Pharma, Actelion, Lundbeck for serving as a speaker and consultant and research support from Genzyme, Shire and Actelion. For conducting of studies on dementia, TD received grants from Novartis and Merz Pharma.
Ethics approval The local ethical board in Münster, Germany.
Provenance and peer review Not commissioned; externally peer reviewed.